Internal Bullwhip Effects (& how to lessen them) – improving time-to-market for scholarly pub – part 1 (of 2)

We played the root beer game in an ops class this semester. I’d heard of bullwhip effects, but it was the first time I’d ever played the game or run a simulation. Interesting game, it made me wonder about implications for scholarly pub.

Basically, the bullwhip effect is a supply chain concept that grew out of research in manufacturing environments: e.g., root beer factories. Relatively recently, (1960s on) researchers found that tiny changes in demand at one end of the pipeline will send destructive shockwaves up the supply chain; the shockwaves increase in amplitude—interfering with and giving rise to other shockwaves—as they travel to the factory, resulting ultimately in complex delays and skyrocketing costs for the organization.

The complex delays and high production costs (overhead) that result are often a mystery to firms suffering these effects. Bullwhip effects can occur externally (originating with the customer) or internally (between divisions or departments). They can also occur in service industry organizations.

My first thoughts ran to books and returns: How much more complicated does “supply” become in a world of “Gone today, here tomorrow” inventory? But again, the bullwhip effect also applies to service industries; e.g., a scholarly press offering publishing services to its authors.

If we look at these “other customers,” authors and manuscripts (mss.), then the supply chain inverts. Rather than printing and warehousing being the “factory” with the customer-facing endpoint being tablets, Amazon, websites, or bookstores, the “factory” would be the production department itself, or the “bookmaking operation” as a whole, and the customer-facing endpoint would be upstream in acquisitions.

Again, organizations can also experience internal bullwhip effects; so from this, the game, and some reading, I had a few questions:

a) Could some scholarly pub houses experience internal bullwhip effects (that might cause fitful delays in delivery of bound books),

b) Could others (that get books out quickly) have procedures in place that coincidentally dampen bullwhip effects, and

c) Could some of the “fixes” developed in other industries shed light on paths to shortened operational timelines and enhanced profitability for u and other scholarly pubs?

Traffic patterns

Proofs for the bullwhip effect are very cool but can be complicated; however, we live through a simple illustration of a linear bullwhip effect whenever we drive through a “backup” on a freeway.

A backup can be caused by a bottleneck or accident. It can also be (and is more frequently) caused by concern over a potential bottleneck or accident or more simply still from merging traffic; i.e., slowing down to observe something unusual or slowing down to avoid another vehicle. In heavy traffic, the car following a car that slows down must slow down a little longer than the car ahead of it, to be sure that the first car resumes speed before continuing on. The car following that car must slow down longer still, and so on down the line. (Each car is over-estimating the need to pause and, in a sense, "asking" for an overestimate from the next car; building in not only a delay but a small multiplier of the delay at each stage.) Fifteen cars back, what began as a 10-second delay for car one has grown to a two and 1/2 minute slowdown for car sixteen, and what's more, the amount added at each stage is increasing: a backup is born.

On the flipside of backups is the world of the “high-speed merge.” A beautiful freeway to drive on, but how to get there? ...and, is it a toll road?

Simple sources and general fixes of bullwhip effects

Per considerable research and many rounds of the root beer game, changes in levels of service or “boom and bust” periods of requests for resources/production cause destructive bullwhip effects. Several general approaches help dampen them: A) keeping service or production levels constant, as much as possible, B) maintaining spare operational capacity (extra room between cars) to absorb the changes that can’t be controlled for — so they don’t translate from one project to the next — and, C) in all cases, sharing information from customer-facing endpoints to supplier-side materials- and services-providers helps managers control against (i.e., lessen) the deleterious impact/s of the shockwaves and bullwhips.

Scholarly pub

Looking within a publishing house, we can see the production department as the “factory” and acquisitions as the “retail store” or customer-facing endpoint; i.e., where the requests of work from the factory begin (e.g., contracts for services, receipts of final mss.). Managing editorial could be lumped in with the “factory” or seen as an intervening “distribution center.” In either case, acquisitions would be a value-added and customer-facing service center leading ultimately to the “factory” beyond.

Together these departments constitute a supply chain of publishing services, resulting in the published work of its customers (authors) landing online and in the warehouse. The receipt of a final ms. and transmittal into the pipeline from acquisitions is the same as an “order” being placed for those services.

(Marketing is part of the publishing services offered; but, delays are already “mature” enough in production, for illustration purposes.)

Treating the “fixes” above in reverse: A) we’d expect any fluctuations in the rate of transmittals from acquisitions to the rest of the house, at the start of the process, to add to scheduling delays and increased overhead costs in the “factory,” per the bullwhip effect, B) we’d expect the absence of spare operational capacity to worsen these effects, and C) any restrictions on the flow of information on future needs for services to send managers through multiple ad hoc rounds of scheduling revision — themselves further consumptive of resources and generative of delays.

Surprises leading to variations in the rates of requests for service (service levels)

Acquisitions, as a “customer-facing endpoint,” contends with considerable variations and surprises in its dealings with authors; e.g., the contracting for and delivery of final mss. 1) Some projects are on a short timeline, sent through peer review and contracted not far in advance of final delivery, as the manuscript is complete. Others are on a longer timeline, as they have yet to be written; i.e., they are contracted on proposal basis. 2) Across both of these categories, some projects arrive late, past their contracted delivery date, while others are on time.

Four general categories of variances and their resulting impact on scheduling are noted below, with projects on short timelines that arrive late offering the most potential surprises ( + + )—and therefore the greatest change in scheduling—and projects on long timelines arriving on time offering the least potential surprises ( – – ).

These fluctuations in scheduling and arrivals can translate to irregularities in transmittals from acquisition to managing editorial and impact scheduling in the production “factory” beyond. That said; some firms may have several steps or other procedures in place that effectively spread ms. arrivals and their transmittals out through the calendar. Such steps or procedures would normalize variations in requests for service and dampen would-be bullwhip effects before they start. Absent these measures, variations in arrivals would translate to variations in requests of later departments in the pipeline.

Divide and accelerate

Isolating groups of projects that characteristically lead to variations in requests of production/requests for resources or present other scheduling challenges (surprises) allows organizations to make managerial decisions to address each group separately, restricting such variations (from getting out of hand) and thereby minimizing costs and maximizing profitability.

There would be many ways to isolate such groups for scholarly publishers; above is just one example. But where would you go from there?

Perhaps (and before addressing operational capacity and information flow) we can look to other industries that manage "controlled chaos" under similar circumstances for inspiration; to see how they cope, survive, and thrive. Many do, in fact, cope, survive and thrive, often with fewer resources and sometimes to quite winning and even profitable effects ...which are categorically better than bullwhip effects.

I'll take a stab at looking for inspiration in one such industry in a followup post.